Abstract
Introduction Chronic back pain is a debilitating condition that significantly impacts quality of life and is a source of healthcare economic burden. Spinal cord stimulation (SCS) is effective in mitigating symptoms, and it has been shown that novel patterns of stimulation can further improve patient experience when appropriately applied. In this study, we compare the effect of different stimulation paradigms on pain scores and objective data collected using digital health tools. The purpose of this study is to determine the feasibility of remotely personalizing and optimizing spinal cord stimulation settings for each patient. Methods Ten subjects, scheduled to undergo SCS treatment for chronic back pain or radiculopathy, will participate in this study. Subjects will perform several study-related tasks on a customized mobile/watch application at baseline and periodically throughout the 4-month study duration. Four stimulation paradigms will be randomly applied in week-long blocks for the initial four weeks after permanent implant: 1) Tonic stimulation (standard and classical high frequency stimulation centered over clinical area of pain, 2) BurstDR: high frequency stimulation performed at subthreshold levels), 3) Mapped Multi-stim: a mixture BurstDR stimulation that is interleaved to move over contacts mapped to the area of pain, and 4) Rostral Multi-stim: a mixture of BurstDR stimulation that is interleaved to move over 4 most rostral contacts. In the following months, we will use Thompson sampling to identify which setting produces the most improvement in pain and recommend it for future use. Results This study will investigate the feasibility of remotely collecting data to classify pain response. Additionally, we will validate the use of this data in optimizing SCS. Lastly, we will explore the relevancy of data (customized app, kinematic, behavior, and EEG) towards phenotyping pain. Conclusions The results from this study will enable ongoing remote optimization of stimulation and comprehensive phenotyping for the purposes of pain treatment. Additionally, this investigation will help identify relevant data streams towards patient-specific optimization that can be collected using ubiquitous sensors to provide a seamless therapeutic experience.